volume | PIER Journals

Abstract

Electromagnetic acoustic transducers (EMATs) have shown broad application prospects in industrial non-destructive testing due to their non-contact and couplant-free operation. However, their low energy conversion efficiency leads to a poor signal-to-noise ratio (SNR), especially under low-power excitation in safety-critical fields such as the petrochemical and nuclear power industries, thereby severely affecting thickness measurement accuracy. To address this challenge, this paper proposes an Adaptive Dual-Attention Fusion Autoencoder (ADFAE) for EMAT echo signal denoising. The ADFAE adopts a dual-path parallel architecture that integrates a multi-scale convolutional autoencoder with channel attention (MCACA) to capture local temporal features and a spatial attention-guided denoising autoencoder (SAGDA) to model global dependencies. Based on the denoised signals, a CNN-BiLSTM network is further employed to directly estimate material thickness. Experimental results demonstrate that the proposed method achieves effective denoising under low SNR conditions, with an average SNR improvement exceeding 23 dB and a mean Peak SNR above 43 dB. Compared with traditional time-of-flight (TOF)-based methods, the proposed ADFAE-CNN-BiLSTM framework significantly improves thickness measurement accuracy, reducing the average relative error to below 0.25%.

1. Aramburu, Arthur Behenck, Joziel Aparecido da Cruz, Amanda Albertin Xavier da Silva, Andrey Pereira Acosta, Larissa Queiroz Minillo, and Rafael de Avila Delucis, "Non-destructive testing techniques for pressure vessels manufactured with polymer composite materials: A systematic review," Measurement, Vol. 246, 116729, 2025.
doi:10.1016/j.measurement.2025.116729 Google Scholar

2. Boughedda, Houssem, Tarik Hacib, Yann Le Bihan, and Hulusi Acikgoz, "Cracks characterization of non-ferromagnetic material using emat probe and PLSR technique," Progress In Electromagnetics Research C, Vol. 103, 199-209, 2020.
doi:10.2528/pierc20013103 Google Scholar

3. Huang, Xiaofei, Yuedong Xie, Fulu Liu, and Lijun Xu, "Electromagnetic acoustic damage imaging using compressive sensing and total focusing method," IEEE Transactions on Instrumentation and Measurement, Vol. 74, 1-9, 2025.
doi:10.1109/tim.2025.3571150 Google Scholar

4. Wang, Chaoqun, Jian Ma, Xue Bai, and Jianwei Chen, "Defect detection and imaging using electromagnetic acoustic transducer with butterfly coil," Review of Scientific Instruments, Vol. 95, No. 6, 064704, 2024.
doi:10.1063/5.0198404 Google Scholar

5. Shankar, Siddharth and Krishnan Balasubramaniam, "Characterising the beam formation of SH waves using double-row staggered Halbach EMAT configurations," NDT & E International, Vol. 121, 102465, 2021.
doi:10.1016/j.ndteint.2021.102465 Google Scholar

6. Si, Dan, Bin Gao, Wei Guo, Yan Yan, G. Y. Tian, and Ying Yin, "Variational mode decomposition linked wavelet method for EMAT denoise with large lift-off effect," NDT & E International, Vol. 107, 102149, 2019.
doi:10.1016/j.ndteint.2019.102149 Google Scholar

7. Cheng, Liang, Maria Kogia, Abbas Mohimi, Vassilios Kappatos, Cem Selcuk, and Tat-Hean Gan, "Crack characterisation using invariable feature extraction in stainless steel specimen used for absorber tubes of CSP applications via EMAT," Renewable Energy, Vol. 101, 771-781, 2017.
doi:10.1016/j.renene.2016.09.036 Google Scholar

8. Mudapaka, Vinay, Thulsiram Gantala, and Krishnan Balasubramaniam, "Development of bulk wave EMAT sensors with enhanced Lorentz force through magnetic field concentration in eddy current regions," Ultrasonics, Vol. 154, 107687, 2025.
doi:10.1016/j.ultras.2025.107687 Google Scholar

9. Dragomiretskiy, Konstantin and Dominique Zosso, "Variational mode decomposition," IEEE Transactions on Signal Processing, Vol. 62, No. 3, 531-544, 2014.
doi:10.1109/tsp.2013.2288675 Google Scholar

10. Guo, Jiangfeng, Ranhong Xie, Yuexiang Wang, Lizhi Xiao, Jianwei Fu, Guowen Jin, and Sihui Luo, "Variational mode decomposition for NMR echo data denoising," IEEE Transactions on Geoscience and Remote Sensing, Vol. 61, 1-14, 2023.
doi:10.1109/tgrs.2023.3237925 Google Scholar

11. Zhou, Jinjie, Dianrui Yu, Xiang Li, Yang Zheng, and Yao Liu, "Reconstructing echoes completely submerged in background noise by a stacked denoising autoencoder method for low-power EMAT testing," Measurement Science and Technology, Vol. 34, No. 12, 125910, Sep. 2023.
doi:10.1088/1361-6501/acf23c Google Scholar

12. Wu, Jing, Yingfeng Yang, Zeyu Lin, Yizhou Lin, Yan Wang, Weiwei Zhang, and Hongwei Ma, "Weak ultrasonic guided wave signal recognition based on one-dimensional convolutional neural network denoising autoencoder and its application to small defect detection in pipelines," Measurement, Vol. 242, 116234, 2025.
doi:10.1016/j.measurement.2024.116234 Google Scholar

13. Wang, Eric Ke, Xun Zhang, and Leyun Pan, "Automatic classification of CAD ECG signals with SDAE and bidirectional long short-term network," IEEE Access, Vol. 7, 182873-182880, 2019.
doi:10.1109/access.2019.2936525 Google Scholar

14. Lu, Haifei, Kedong Zhou, and Lei He, "Bearing fault vibration signal denoising based on adaptive denoising autoencoder," Electronics, Vol. 13, No. 12, 2403, 2024.
doi:10.3390/electronics13122403 Google Scholar

15. Hu, Jie, Li Shen, and Gang Sun, "Squeeze-and-excitation networks," Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 7132-7141, Salt Lake City, UT, USA, 2018.

16. Han, Shuzhen, Shengke Sun, Zhanshan Zhao, Ziqian Luan, and Pingjuan Niu, "Deep residual multiscale convolutional neural network with attention mechanism for bearing fault diagnosis under strong noise environment," IEEE Sensors Journal, Vol. 24, No. 6, 9073-9081, 2024.
doi:10.1109/jsen.2023.3345400 Google Scholar

17. Zhang, Y., S. L. Huang, S. Wang, and W. Zhao, "Time-frequency energy density precipitation method for time-of-flight extraction of narrowband lamb wave detection signals," Review of Scientific Instruments, Vol. 87, No. 5, 054702, 2016.
doi:10.1063/1.4948426 Google Scholar

18. Liu, Wangyu, Dongxun Li, and Weigui Xie, "A novel time-of-flight difference determination method for ultrasonic thickness measurement with ultrasonic echo onset point detection," Applied Acoustics, Vol. 233, 110605, 2025.
doi:10.1016/j.apacoust.2025.110605 Google Scholar

19. Dive, Vikas Pralhad, Sandesh Solepatil, Amol Mali, Sunil Dambhare, and Aniket Kolekar, "Investigation of the effect of grain size on longitudinal critically refracted ultrasonic wave time-of-flight and velocity of propagation in material," Advances in Science and Technology Research Journal, Vol. 19, No. 6, 391-402, 2025.
doi:10.12913/22998624/203515 Google Scholar

20. Wan, Tao, Takashi Naoe, Takashi Wakui, Masatoshi Futakawa, Hironari Obayashi, and Toshinobu Sasa, "Effects of grain size on ultrasonic attenuation in type 316L stainless steel," Materials, Vol. 10, No. 7, 753, 2017.
doi:10.3390/ma10070753 Google Scholar

21. Zhou, Ya and Xiaobo Jiao, "Intelligent analysis system for signal processing tasks based on LSTM recurrent neural network algorithm," Neural Computing and Applications, Vol. 34, No. 15, 12257-12269, 2022.
doi:10.1007/s00521-021-06478-6 Google Scholar

22. Shang, Li, Zi Zhang, Fujian Tang, Qi Cao, Hong Pan, and Zhibin Lin, "CNN-LSTM hybrid model to promote signal processing of ultrasonic guided lamb waves for damage detection in metallic pipelines," Sensors, Vol. 23, No. 16, 7059, 2023.
doi:10.3390/s23167059 Google Scholar

23. Yang, Lehui, Jinling Zhao, Le Chang, Jianping Zhao, and Jingwei Cheng, "Evaluating the thickness of thin-walled structures from ultrasound signals based on CNN-BILSTM model," Nondestructive Testing and Evaluation, 1-19, 2025.
doi:10.1080/10589759.2025.2507763 Google Scholar

24. Kubinyi, Michal, Ondrej Kreibich, Jan Neuzil, and Radislav Smid, "EMAT noise suppression using information fusion in stationary wavelet packets," IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control, Vol. 58, No. 5, 1027-1036, 2011.
doi:10.1109/tuffc.2011.1903 Google Scholar

25. Lu, Zhengyang and Ying Chen, "Single image super-resolution based on a modified U-Net with mixed gradient loss," Signal, Image and Video Processing, Vol. 16, No. 5, 1143-1151, 2022.
doi:10.1007/s11760-021-02063-5 Google Scholar

26. Maev, R. G., H. Shao, and E. Yu. Maeva, "Thickness measurement of a curved multilayered polymer system by using an ultrasonic pulse-echo method," Materials Characterization, Vol. 41, No. 2-3, 97-105, 1998.
doi:10.1016/s1044-5803(98)00030-8 Google Scholar

27. Fan, Caizi, Yongchao Zhang, Hui Ma, and Zeyu Ma, "A novel deep denoising model integrating transformer and time-frequency loss for gearbox fault diagnosis," Advanced Engineering Informatics, Vol. 66, 103400, 2025.
doi:10.1016/j.aei.2025.103400 Google Scholar

28. Wang, Pei, Haizhen Tao, Jingshuai Qi, and Peng Li, "Machining quality prediction of complex thin-walled parts using multi-task dual domain adaptive deep transfer learning," Advanced Engineering Informatics, Vol. 62, 102640, 2024.
doi:10.1016/j.aei.2024.102640 Google Scholar

29. Lomoio, Ugo, Pierangelo Veltri, Pietro Hiram Guzzi, and Pietro Liò, "Design and use of a denoising convolutional autoencoder for reconstructing electrocardiogram signals at super resolution," Artificial Intelligence in Medicine, Vol. 160, 103058, 2025.
doi:10.1016/j.artmed.2024.103058 Google Scholar

30. Cai, Jianxian, Li Wang, Jiangshan Zheng, Zhijun Duan, Ling Li, and Ning Chen, "Denoising method for seismic co-band noise based on a U-Net network combined with a residual dense block," Applied Sciences, Vol. 13, No. 3, 1324, 2023.
doi:10.3390/app13031324 Google Scholar

Dr. Pan Guo

Dr. Zhuorui Zhang

Dr. Qiuyan Zhong